A Simulation Method of Establishing Fatigue Life Distribution

Abstract

This paper presents a Monte Carlo simulation method for fatigue failure, by which the randomness of two material properties as well as that of the applied load can be incorporated into a stochastic model using an appropriate failure criterion to predict the statistical characteristics of fatigue life under constant and random amplitude cyclic loading conditions. In this technique, both the endurance limit S e and the fatigue strength coefficient S f ′ are treated as stochastic variables. The combined effect of the randomness of S e , S f ′ , and the applied stress on the statistical characteristics of fatigue lives is predicted analytically using digital simulation of fatique tests. The life distributions and their statistical characteristics are found to be in good agreement with those obtained from analyzing the experimental results, indicating that the proposed technique and the underlying assumptions and hypotheses are adequate. The suggested method is believed to be an effective, fast, and easy-to-use design tool which is suitable for use on electronic computers. It is ideal for parametric studies compared with the costly and time-consuming laboratory fatigue tests. Minimum experimental data are needed as a basis for the analysis. New results are presented which show the effect of the randomness of the loads and material properties on the randomness of fatigue life distribution.